Skip to content
orbyd

Dossier · AMBA · Dormant

AMBA · Ambarella, Inc. · Stock research

MEDIUM Cyclical recovery physical-ai-robotics ai-chips-memory

Last analysed ·

Current thesis

Physical-AI pure-play narrative accelerating: Rosenblatt named AMBA a top H2-2026 pick (2026-07-01, PT $120) and the stock ripped ~28-32% off the June flush, reclaiming $72 and breaking to ~$80. Hanwha's >$800M deal underwrites demand, but entry at ~$77 chases a vertical move into insider selling, with no earnings catalyst until ~late-August.

Invalidation trigger

A weekly close below $72 forfeits the post-Rosenblatt breakout shelf and returns price to the broken June range; below the $63.50 June swing low resumes the breakdown. Secondary: Q2 FY27 (~late Aug) revenue under the $105M guide floor, a Hanwha ramp delay, or the physical-AI theme flipping to SATURATED.

Thesis status

Open commitment scored if the trigger above fires How this is scored →

Latest analysis and events for AMBA —

As of 2026-07-11, orbyd's latest analysis for Ambarella, Inc. (AMBA): Physical-AI pure-play narrative accelerating: Rosenblatt named AMBA a top H2-2026 pick (2026-07-01, PT $120) and the stock ripped ~28-32% off the June flush, reclaiming $72 and breaking to ~$80. Hanwha's >$800M deal underwrites demand, but entry at ~$77 chases a vertical move into insider selling, with no earnings catalyst until ~late-August.

Invalidation trigger: A weekly close below $72 forfeits the post-Rosenblatt breakout shelf and returns price to the broken June range; below the $63.50 June swing low resumes the breakdown. Secondary: Q2 FY27 (~late Aug) revenue under the $105M guide floor, a Hanwha ramp delay, or the physical-AI theme flipping to SATURATED.

Current Thesis

The shakeout thesis resolved to the upside. After the 2026-06-05 flush to $63.52 — that broke the $68 base, Ambarella rebuilt a higher low, reclaimed the $72 weekly line the structure had to clear, and then broke out hard: a ~28–32% single-session surge around 2026-06-30 on Rosenblatt naming AMBA a top pick for H2 2026 and a "physical-AI pure play" (Buy, PT lifted to $120 from $115). Price ran to a high near ~$80 and now consolidates ~$77.30 (2026-07-10 close). The edge-AI/physical-AI silicon narrative is accelerating and sell-side is chasing it up rather than fading it. What keeps a fresh entry here honest: the buy point is now ~28% off the June low into a vertical move, insiders keep selling into the strength, and there is no company-specific catalyst until the ~late-August Q2 print so the story rides on theme momentum and the Hanwha ramp, not a dated event.

Bullish and bearish views on Ambarella, Inc.

The model's bull view on Ambarella, Inc. (AMBA), in brief: Rosenblatt top-pick catalyst (2026-07-01): analyst Kevin Cassidy named AMBA a top H2-2026 idea and a "physical-AI pure play," reiterating Buy and raising the target to $120 from $115, tied to low-power AI vision processors in surveillance, robotics, drones, industrial automation… The bear view: Chasing a vertical move: a fresh entry at ~$77 is buying the day after a ~28–32% single-session blowoff, up ~28% off the June low, on a small-cap that historically prints 18–22% single-session air pockets. Both cases follow in full.

Bull Case

  • Rosenblatt top-pick catalyst (2026-07-01): analyst Kevin Cassidy named AMBA a top H2-2026 idea and a "physical-AI pure play," reiterating Buy and raising the target to $120 from $115, tied to low-power AI vision processors in surveillance, robotics, drones, industrial automation and autonomous systems. This drove the ~28–32% single-session rip on 2026-06-30.
  • Theme accelerating at the sector level: semiconductor/edge-AI leadership is broad, and Ambarella is being repositioned by the Street as an inference-at-the-sensor supplier as compute shifts from cloud to edge devices a durable multi-year demand vector rather than a single quarter.
  • Hanwha agreement (2026-05-28): >$800M over 10+ years to co-develop and source edge-AI SoCs/software across video security, robotics, industrial automation and life sciences. Hanwha Vision is a ~15-year customer; management guided "some positive revenue impact next year," converting lumpy orders toward a contracted base.
  • Automotive at a record and robotics design wins (Q1 FY27 call, 2026-05-28): auto revenue hit an all-time high on AI adoption in commercial-vehicle telematics/safety, alongside 15+ robotics design wins the longest-cycle, highest-margin end markets.
  • Room to the consensus target: with spot ~$77 versus a consensus PT clustered ~$95–100 (range $80–$120) and a 52-week high of $96.69, the sell-side still models upside even after the pop.

Bear Case

  • Chasing a vertical move: a fresh entry at ~$77 is buying the day after a ~28–32% single-session blowoff, up ~28% off the June low, on a small-cap that historically prints 18–22% single-session air pockets. The clean risk/reward buy was the $63–66 flush, not the post-catalyst extension.
  • continuation of the June distribution (CEO, CFO, COO and an SVP all sold into the mid-to-high $60s). Management is trimming into the rally, not adding.
  • Growth is modest against the multiple: Q1 FY27 revenue of $100.4M is +16.9% YoY; the Q2 guide of $105–111M is in-line, not a beat-and-raise. GAAP net loss was −$18.1M (58.4% GM), so on GAAP there is no earnings support under the stock if sentiment cools.
  • No dated catalyst for ~6 weeks: the next binary event is the Q2 FY27 print (~late August). Between now and then the tape has to hold a rich valuation on theme flow alone, and a stalled physical-AI narrative removes the bid.
  • High-beta whippy small cap: ~$3.39B market cap, beta ~2.15. The name moves roughly 2x the tape in both directions; position size has to respect the air-pocket risk.

Setup & Price Structure

  • Spot ~$77.30 (2026-07-10 close, −1.38% on the day); previous close $78.38, after-hours ~$77.50. Market cap $3.39B, GAAP-unprofitable (trailing P/E n/a; Street forward P/E ~85). 52-week range $48.30–$96.69.
  • Breakout confirmed off the June flush: the $63.52 low (2026-06-05) held as a higher low, $72 was reclaimed, and the 2026-06-30 catalyst broke price to ~$80. The $72 shelf is now the line the breakout must defend; the ~$76–80 zone is the fresh consolidation base being built.
  • Position vs. targets: spot sits ~15–20% below the $96.69 52-week high and just under the ~$95–100 consensus target closer to fair value than it was a month ago, which trims the margin of safety on a chase.
  • a pullback toward $72–74 would be a healthier re-entry than the current print.

Catalyst Calendar (next 30 days)

  • No scheduled binary event in the 2026-07-11 → 2026-08-10 window. Q2 FY27 covers the quarter ended 2026-07-31; the print is expected ~late August (est. ~2026-08-27; prior-year Q2 was reported 2025-08-28) outside the 30-day window. Revenue guide $105–111M.
  • Theme/sector tape: physical-AI and edge-AI headline flow and semiconductor-basket leadership are the day-to-day driver absent a company event.

Elapsed catalysts

  • Sell-side follow-through: watch for additional target revisions or new coverage after the Rosenblatt top-pick call (2026-07-01) clustered upgrades would confirm the narrative is still accelerating. _(passed 18d ago)_
  • Insider Form 4 flow: continued open-market sales (latest 5,958 sh @ $78.66, 2026-07-09) are a running tell on how management values the rally. _(passed 10d ago)_

What Would Change Our Mind

  • Bull invalidation: a weekly close below $72 forfeits the post-Rosenblatt breakout shelf and puts price back inside the broken June range; a close below the $63.50 June swing low resumes the outright breakdown and negates the shakeout read.
  • Fundamental break: Q2 FY27 (~late August) revenue under the $105M guide floor, any Hanwha ramp delay or walk-back, or the physical-AI theme flipping to SATURATED (mainstream CNBC-cover mania with no fresh design wins) removes the leg the current multiple rests on.
  • Bull confirmation: holding the $76–80 base and a weekly close back above ~$82–85 reopens a path to the $96.69 high and keeps the theme-leader bid intact.

Correlation Notes

  • High-beta (~2.15) small-cap semiconductor; trades with the SOX/semiconductor complex and the broader edge/physical-AI basket that Rosenblatt's framing groups it into. Sector risk-off drags it ~2x the index.
  • Idiosyncratic drivers dominate on catalyst days: the Hanwha ramp, robotics/automotive design-win cadence, and insider Form 4 flow move the name independent of the tape.
  • Whippy execution profile 18–22% single-session moves in both directions in recent history

Notes

  • FISCAL CALENDAR: this was Q1 FY2027 (period ended 2026-04-30), printed 2026-05-28 NOT 'Q1 FY2026' as the stale prior dossier implied. Next print ~late Aug 2026 (Q2 FY27).
  • Q1 FY27: rev $100.4M +16.9% YoY, adj EPS $0.11 beat $0.10, GAAP net loss -$18.1M, GAAP GM 58.4%. Auto at all-time record, 15+ robotics design wins.
  • Hanwha deal (2026-05-28): >$800M over 10+ yrs, edge-AI across security/robotics/industrial/life-sciences. THIS, not the numbers, drove the +9.8% pop.
  • Post-print PTs: Rosenblatt Buy $120, Susquehanna $110, BofA Neutral $96 all 2026-05-29. Spot ~$73 well below all.
  • Earnings blackout: no near-term earnings catalyst print just passed. This is a no-binary-risk window; thesis rides on theme + Hanwha ramp.
  • Small-cap (~$3B), high-beta, whippy Tickeron flagged a -20.9% single-session drop in recent history. Size accordingly.
  • FISCAL CALENDAR: last print was Q1 FY2027 (period ended 2026-04-30), reported 2026-05-28. Next print ~late Aug 2026 (Q2 FY27). No earnings in the next 30d no binary risk and no print-driven upside near term.
  • STRUCTURE BROKEN 2026-06-05: weekly close $63.52, −11.84% (−$8.53), through the $68 base and the $68.80 Apr-30 pre-print shelf. Prior re-rating invalidation has fired; treat as busted, not a dip. Do not anchor to the ~$96 consensus PT.
  • June 5 drop had no single clean catalyst continuation of valuation rejection (SA bear piece 2026-06-03), Summit Insights downgrade to Hold on 2H26 supply-chain/demand risk, and CEO/CFO insider selling into a high-beta (2.15) small cap (~$2.79B). NOT patent-related (the GoPro/Insta360 ITC matter resolved Feb 2026 with limited impact).
  • Hanwha deal (2026-05-28): >$800M over 10+ yrs across security/robotics/industrial/life-sciences; mgmt guided 'some positive revenue impact next year,' not an immediate step-up. Hanwha Vision a ~15-yr customer.
  • Q1 FY27: rev $100.4M +16.9% YoY, adj EPS $0.11 beat $0.10, GAAP net loss −$18.1M, GAAP GM 58.4%; auto record, 15+ robotics design wins; Q2 guide $105–111M vs $107.4M est. FY26 rev $390.7M, IoT +~50% YoY, 46M+ edge-AI units installed, 12 edge-AI SoCs.
  • Post-print PTs (2026-05-29): Rosenblatt Buy $120, Susquehanna Positive $110, BofA Neutral $96, Roth Neutral $90 (from $65). Summit Insights downgraded to Hold. Consensus ~Moderate Buy (1 strong buy / 7 buy / 5 hold / 1 sell).
  • Customer concentration: Arashi Vision (Insta360) is the largest customer; structural single-name risk even though the GoPro ITC overhang is resolved. High beta (2.15) size any probe small and only on a confirmed re-base above $72.
  • FISCAL CALENDAR: last print was Q1 FY2027 (period ended 2026-04-30), reported 2026-05-28. Next print Q2 FY27 ~late Aug 2026. No earnings in next 30d no binary print risk through ~2026-07-21.
  • STRUCTURE REPAIRING 2026-06-20: after the 2026-06-05 weekly close $63.52 (-11.84%) broke the $68 base, price rebuilt a higher-low sequence ($63.52 → ~$65.88 6/17 → $68.19 6/18 → ~$70.01 6/20, +6.21%) and is pressing the $72 weekly-reclaim line. June flush now reads as a likely shakeout, not a trend-down but unconfirmed below $72.
  • KEY LEVEL: $72 weekly close = confirmation gate (turns the June flush into a confirmed bear-trap / cleaner long). ~$63.50 (June 5 low) = breakdown resumes. $48.30 = 52-week structural floor.
  • Q1 FY27: rev $100.4M +16.87% YoY, adj EPS $0.11 beat $0.10, GAAP net loss -$18.1M, GAAP GM 58.4%; auto record, 15+ robotics design wins; Q2 guide $105-111M (in-line, not beat-and-raise).
  • Post-print PTs (raised into the dip): Susquehanna $110 (from $90, Positive), Rosenblatt $120 Buy, BofA $96 (from $72, Neutral), Roth $90 (from $65, Neutral). Consensus ~$93.75, range $80-$120, Buy. Spot ~$70 — below all.
  • Continued insider selling into the bounce (~$67-68): CEO Wang 6,204 sh ($421k), COO Lee 2,951 ($200k), CFO Young 3,186 ($216k), SVP Ju 2,017 ($137k). Distribution, not accumulation.
  • GoPro won a patent case vs Insta360 (late June 2026) fresh overhang on AMBA's action-camera partner ecosystem; distinct from the Feb 2026 ITC matter (resolved, limited impact).
  • Inaugural quarterly Industry Analyst Briefing held 2026-06-04 (Cooper Developer Platform, Developer Zone); quarterly cadence → next ~Sept. Narrative event, not a numbers print.
  • Small-cap (~$3.07B), high-beta (2.15), whippy documented -18% to -22% single-session drops. Size as a probe.
  • FISCAL CALENDAR: last print was Q1 FY2027 (period ended 2026-04-30), reported 2026-05-28. Next print is Q2 FY27 (period ended 2026-07-31), expected ~late Aug 2026 (est. ~2026-08-27; prior-year Q2 reported 2025-08-28). No earnings in the next 30d no binary risk near term.
  • NARRATIVE SHIFT 2026-06-30/07-01: Rosenblatt (Kevin Cassidy) named AMBA a top H2-2026 pick and 'physical-AI pure play', Buy, PT raised to $120 from $115. Stock ripped ~28-32% in a single session. This is the acceleration event that flipped the June breakdown into a confirmed shakeout+breakout.
  • Q1 FY27 (reported 2026-05-28): rev $100.4M +16.9% YoY, adj EPS $0.11 beat $0.10, GAAP net loss -$18.1M, GAAP GM 58.4%; automotive at all-time record, 15+ robotics design wins. Q2 guide $105-111M (in-line).
  • Hanwha deal (2026-05-28): >$800M over 10+ yrs across security/robotics/industrial/life-sciences; Hanwha Vision a ~15-yr customer; mgmt guided 'some positive revenue impact next year' a ramp, not an immediate step-up.
  • INSIDER SELLING PERSISTS: VP Yun-Lung Chen sold 5,958 sh @ $78.66 on 2026-07-09 (~$469K), 60,557 remaining. Continuation of June selling by CEO/CFO/COO/SVP into the mid-to-high $60s. Distribution into strength, not accumulation.
  • Small-cap (~$3.39B), beta ~2.15, whippy 18-22% single-session air pockets both directions. Size accordingly. GAAP-unprofitable, trailing P/E n/a, Street forward P/E ~85.
  • STRUCTURE: $63.52 (2026-06-05) is the June swing low; $72 is the reclaimed breakout shelf now acting as support; $76-80 is the post-catalyst consolidation base. 52-wk range $48.30-$96.69; consensus PT ~$95-100 (range $80-$120).

Related · shared themes

SIMO

Silicon Motion Technology Corporation

NAND-controller toll-booth on the steepest flash shortage in ~15 years: TrendForce H1 2026 contract +>100% cumulative, H2 still rising with no capacity adds. Q1 record +105% YoY, sell-side chasing to $400/$450. Fundamental leg ACCELERATING, but the tape is digesting a 4x into insider selling the 2026-07-29 Q2 print is the next binary.

MEDIUM

AMD

Advanced Micro Devices, Inc.

MI450/Helios rack-scale plus EPYC agentic-CPU narrative drove ATHs near $561.80 (7/1), but the July target cluster ($600 BNP, $620 BofA, $725 high) is arriving into a falling tape 20-day MA lost 7/8, chip-led Nasdaq selloff 7/17, sentiment into Fear. Sell-side catch-up after the move, with the 8/4 print as the binary.

LOW

HPE

Hewlett Packard Enterprise Company

AI-server and enterprise-networking pivot is real at the theme level, but HPE's own chart broke: the record +25% Q2 FY26 blowoff (2026-06-02, ATH $64.25) fully round-tripped, the ~$48 pre-print shelf failed (last observed close $43.71 on 6/26), and nothing is dated until the Q3 print in early September. A fresh buy here catches the breakdown, not a base.

LOW

MU

Micron Technology, Inc.

HBM scarcity intact SK Hynix sees shortage into the next decade, Qualcomm/Hyundai auto deals signed 2026-07-16, US push to ban Chinese memory but the tape broke: SOXX −18.6% in July, worst month since 2008, and the $854.50 shelf failed. At 6.2x forward the setup is cycle-timing, not momentum; no higher low yet.

LOW

See also · stocks to watch

AI compute stocks to watch